Project Summary / Abstract The objective of the proposed research is to better understand how peripheral blood microbiome patterns are associated with knee osteoarthritis (OA). Our laboratory has previously examined in detail the biomarker potential of peripheral blood epigenetic changes both to predict future rapid radiographic/pain progression in patients who already have early knee OA, as well as to detect preclinical OA (e.g. patients who do not have radiographic OA but will develop it in the near future). Thus, we have access to a large number (>1800) of peripheral blood DNA specimens that can be used for additional analyses. Furthermore, our laboratory and others have shown the importance of the gut microbiome to OA susceptibility and progression, and we have recently generated the first data regarding the presence of an articular (joint) microbiome. Presumably this joint microbiome develops due to hematogenous spread of bacteria/bacterial products from the gut. In the current study, we will look for the presence of these bacteria/bacterial DNA within peripheral blood samples in the context of OA. Our first Aim is to determine whether differences in baseline peripheral blood bacterial DNA patterns are predictive of future radiographic and/or pain progression (within the next 24 months) in early knee OA patients, and we will develop machine-learning biomarker models to discriminate future progressors from nonprogressors, using DNA samples from the Osteoarthritis Initiative (OAI), the Multicenter Osteoarthritis Study (MOST), and the Systematic Oklahoma Osteoarthritis iNflammation and Epigenetics Research (SOONER) cohort. In our second Aim, we will determine whether differences in baseline peripheral blood bacterial DNA patterns can be used to predict preclinical OA among healthy patients from the OAI who went on to develop radiographic knee OA within 12-96 months. The proposed work is important, as we do not have a full understanding of how various microbiome niches influence OA risk, nor do we have robust clinically-available biomarkers to distinguish future OA progressors or detect preclinical disease. Our work is quite innovative in its hypothesis of a gut:blood:joint microbiome in OA pathogenesis, as well as the use of cutting-edge deep sequencing of microbial DNA and machine learning methods to develop discriminatory models. Success in our proposal may open a new avenue for OA microbiome research and may offer a novel treatment strategy for OA.